Winboard Forum

by **Vladimir Medvedev** » 12 Feb 2007, 12:46

http://greko.chess.googlepages.com/greko-5.2-win.zip
http://greko.chess.googlepages.com/greko-5.2-linux.zip

Version 5.2
12-Feb-2007
-----------

(fix) Main hash table size is not divided by 2 anymore.

(new) Added futility pruning, in a very simple and straightforward form.

(chg) Search: tuned some conditions for LMR and Nullmove.
(chg) Eval: King safety value scaled up a little.

Overall improvement in strength is 50...70 points for blitz games.

by **Tony Thomas** » 12 Feb 2007, 16:10

Vladimir Medvedev wrote:http://greko.chess.googlepages.com/greko-5.2-win.zip
http://greko.chess.googlepages.com/greko-5.2-linux.zip

Version 5.2
12-Feb-2007
-----------

(fix) Main hash table size is not divided by 2 anymore.

(new) Added futility ptuning, in a very simple and straightforward form.

(chg) Search: tuned some conditions for LMR and Nullmove.
(chg) Eval: King safety value scaled up a little.

Overall improvement in strength is 50...70 points for blitz games.

Your engine seem to be a threat to Rybka at this rate.

by **Vladimir Medvedev** » 12 Feb 2007, 16:14

Tony Thomas wrote:Your engine seem to be a threat to Rybka at this rate.

Simple plot in Excel shows me that GreKo is somewhere near it's asymptotic saturation, at the level of maybe 2400...

by **Tony Thomas** » 12 Feb 2007, 20:46

Vladimir Medvedev wrote:
Tony Thomas wrote:Your engine seem to be a threat to Rybka at this rate.

Simple plot in Excel shows me that GreKo is somewhere near it's asymptotic saturation, at the level of maybe 2400...

It was already around 2350 with the previous version, now with this improvement you probably broke the 2400 barrier. You are the programmer, you know about your program better than me.

by **Vladimir Medvedev** » 13 Feb 2007, 09:06

Maybe 50...70 points was overestimation. In my toy rating pool the difference is smaller.
http://greko.chess.googlepages.com/openrating

Code: Select all: Program Elo + - Games Score Av.Op. Draws 1 Toga II 1.2.1a : 2731 89 84 84 86.3 % 2411 17.9 % 2 Fruit 2.1 : 2700 98 92 84 83.9 % 2413 8.3 % 3 Scorpio 1.8 : 2600 76 73 84 73.8 % 2420 19.0 % 4 SlowChessBlitz 2.96 : 2558 75 72 84 68.5 % 2423 15.5 % 5 Delfi 5.0 : 2554 72 70 84 67.9 % 2424 19.0 % 6 Glaurung 1.2.1 : 2549 73 71 84 67.3 % 2424 17.9 % 7 Crafty 20.14 : 2545 77 74 84 66.7 % 2424 9.5 % 8 Phalanx XXII : 2425 69 69 84 48.8 % 2433 16.7 % 9 GreKo 5.2 : 2354 72 73 78 37.2 % 2445 17.9 % 10 EXchess 5.0.1 beta : 2350 69 71 84 37.5 % 2438 17.9 % 11 GreKo 5.1 : 2330 76 78 78 34.0 % 2445 11.5 % 12 GreKo 2.817 : 2320 74 77 84 33.3 % 2440 9.5 % 13 GNUChess 4 TM : 2307 76 79 84 31.5 % 2441 8.3 % 14 HoiChess 0.6.1 : 1985 149 121 84 6.0 % 2464 0.0 % 15 TSCP 1.81 : 1967 153 101 84 5.4 % 2466 1.2 %

by **Tony Thomas** » 13 Feb 2007, 09:35

I had chosen Greko as a sparring partner for Romichess, but I changed my mind later because Greko was getting stronger at an alarming rate, and I was looking for two engines weaker than Romi. It looks to me after around 180 games that the current version of Romi is way too strong for the weaker engines that I picked, so I am going to let Greko feel the heat from Romi. :mrgreen:

by **Vladimir Medvedev** » 23 Feb 2007, 14:41

New update:

http://greko.chess.googlepages.com/greko-5.2.5-win.zip
http://greko.chess.googlepages.com/grek ... source.zip

by **Tony Thomas** » 23 Feb 2007, 21:20

You release versions so fast that I barely have time to play with them. Here is my private database that I created to test future versions of Romichess. Its not finished yet, but its almost done. Note that Zappa 1.1's rating is fixed at 2600. Provided that it is at least a 2700 engine, add about 100 points, you get the realistic rating of some of the engines. :mrgreen:

2/23/2007 3:19:50 PM :

Program Elo + - Games Score Av.Op. Draws

Code: Select all: 1 Zappa 1.1 : 2600 36 35 360 71.5 % 2440 16.4 % 2 Arasan 9.5 : 2558 35 34 360 66.0 % 2443 15.3 % 3 Francesca MAD 0.13 : 2550 34 33 360 64.9 % 2444 19.2 % 4 Danasah 2.85 : 2498 32 32 360 57.1 % 2448 20.8 % 5 Phalanx Reborn : 2475 33 33 360 53.6 % 2450 16.7 % 6 RomiChessDK2 : 2475 33 33 360 53.6 % 2450 16.1 % 7 Delphil 1.6c : 2455 37 37 284 48.2 % 2467 18.3 % 8 NanoSzachy 2.7 : 2433 41 41 233 40.1 % 2502 19.3 % 9 Djinn 0.925x : 2417 39 39 270 42.0 % 2473 15.2 % 10 Horizon 4.3 : 2415 44 44 210 36.4 % 2512 17.6 % 11 Zeus 1.28 : 2409 33 33 360 43.3 % 2456 17.2 % 12 GreKo 5.2 : 2347 45 46 219 28.3 % 2509 16.4 % 13 Lime 6.3 : 2214 50 51 270 16.5 % 2495 10.7 %

by **Guenther Simon** » 23 Feb 2007, 21:30

Tony Thomas wrote:You release versions so fast that I barely have time to play with them. Here is my private database that I created to test future versions of Romichess. Its not finished yet, but its almost done. Note that Zappa 1.1's rating is fixed at 2600. Provided that it is at least a 2700 engine, add about 100 points, you get the realistic rating of some of the engines.

2/23/2007 3:19:50 PM :

Program Elo + - Games Score Av.Op. Draws

Code: Select all
1 Zappa 1.1 : 2600 36 35 360 71.5 % 2440 16.4 % 2 Arasan 9.5 : 2558 35 34 360 66.0 % 2443 15.3 % 3 Francesca MAD 0.13 : 2550 34 33 360 64.9 % 2444 19.2 % 4 Danasah 2.85 : 2498 32 32 360 57.1 % 2448 20.8 % 5 Phalanx Reborn : 2475 33 33 360 53.6 % 2450 16.7 % 6 RomiChessDK2 : 2475 33 33 360 53.6 % 2450 16.1 % 7 Delphil 1.6c : 2455 37 37 284 48.2 % 2467 18.3 % 8 NanoSzachy 2.7 : 2433 41 41 233 40.1 % 2502 19.3 % 9 Djinn 0.925x : 2417 39 39 270 42.0 % 2473 15.2 % 10 Horizon 4.3 : 2415 44 44 210 36.4 % 2512 17.6 % 11 Zeus 1.28 : 2409 33 33 360 43.3 % 2456 17.2 % 12 GreKo 5.2 : 2347 45 46 219 28.3 % 2509 16.4 % 13 Lime 6.3 : 2214 50 51 270 16.5 % 2495 10.7 %

I know no list in which Zappa on a single machine is anywhere
near 2700 in relation to what you list here.

Guenther

by **Graham Banks** » 24 Feb 2007, 10:11

Guenther Simon wrote:I know no list in which Zappa on a single machine is anywhere
near 2700 in relation to what you list here.

Guenther

Code: Select all: CCRL 40/40 Rating List - 32-bit single-CPU engines 48877 games played by 254 programs, run by 13 testers Ponder off, General books (up to 12 moves), 3-4-5 piece EGTB Time control: Equivalent to 40 moves in 40 minutes on Athlon 64 X2 4600+ (2.4 GHz) Computed on February 24, 2007 with Bayeselo based on 48'877 games Tested by CCRL team, 2005-2007, http://computerchess.org.uk/ccrl/4040/ Rank Engine ELO + - Score AvOp Games 1 Rybka 2.2 32-bit 2995 +21 -21 72.2% -153.1 819 Rybka 1.2 32-bit 2977 +18 -18 70.7% -149.3 1075 Rybka 2.1 32-bit 2970 +23 -23 69.5% -130.6 646 Rybka 2.1 32-bit (C+1 OSO) 2966 +31 -30 72.3% -160.5 410 Rybka 1.1 32-bit 2960 +18 -17 69.7% -138.5 1162 Rybka 2.3 32-bit 2940 +85 -83 59.0% -51.0 39 Zap!Chess Zanzibar 32-bit 2889 +51 -51 54.1% -26.4 109 Rybka 1.0 32-bit 2884 +16 -16 60.7% -72.7 1240 2 Hiarcs 11 2876 +29 -28 54.9% -29.4 379 3 Fritz 10 2873 +33 -33 55.6% -38.4 296 4 Shredder 10 2872 +16 -15 60.7% -72.1 1392 5 Loop 10.32f 2871 +19 -19 51.0% -5.8 841 Hiarcs X50 Hypermodern 2864 +28 -28 50.2% -1.7 419 6 Toga II 1.2.1a 2862 +13 -13 49.6% +0.6 2010 Loop 12.32 2856 +37 -36 56.6% -44.5 234 7-8 Naum 2.1 32-bit 2850 +32 -32 54.4% -32.3 306 7-8 Spike 1.2 Turin 2850 +13 -13 51.1% -8.5 2002 Toga II 1.2 32-bit 2846 +32 -32 55.5% -38.1 309 Hiarcs X50 2845 +17 -18 45.8% +30.5 1114 Fritz 9 2842 +13 -13 50.0% +0.8 2111 Hiarcs X54 Hypermodern 32-bit 2841 +40 -40 49.3% +5.0 203 Fruit 2.2.1 2840 +13 -13 52.8% -17.7 1973 Hiarcs 10 2839 +15 -15 50.1% +0.8 1494 9 Junior 10 2839 +18 -18 50.3% -2.5 1006 Hiarcs X54 32-bit 2828 +36 -36 45.3% +34.2 247 Toga II 1.1a 2826 +18 -18 55.2% -34.3 1017 Chess Tiger 2007 Gambit 2823 +349 -239 90.0% -240.8 5 Shredder 9 2823 +17 -17 55.7% -38.3 1162 Junior 10.1 2821 +93 -93 51.3% -9.5 39 Hiarcs 10 Hypermodern 2818 +19 -19 51.7% -8.4 889 Hiarcs 11.1 2816 +170 -170 50.0% +13.9 9 Scorpio 1.82 2814 +183 -191 43.8% +36.0 8 10 Ktulu 8.0 2803 +17 -17 43.2% +45.3 1163 Scorpio 1.91 2801 +61 -60 60.0% -63.2 90 Fritz 8 Bilbao 2800 +27 -27 44.9% +37.1 442 Naum 2.0 32-bit 2800 +18 -18 51.0% -4.1 959 11 Chess Tiger 2007 2795 +18 -18 46.5% +23.3 961 Spike 1.1 2786 +15 -15 49.0% +6.2 1469 Scorpio 1.84 2784 +53 -53 56.7% -71.4 141 Scorpio 1.9 2783 +55 -55 50.0% -6.7 114 12 Zap!Chess Paderborn 32-bit 2781 +21 -21 45.2% +34.5 732 Junior 9 2780 +16 -16 46.6% +21.7 1237 Glaurung 1.2.1 Avalanche 32-bit 2772 +48 -48 52.4% -17.7 147 Zap!Chess Paderborn SE 32-bit 2772 +36 -36 45.7% +28.4 245 Fruit 2.1 2770 +49 -48 54.7% -42.0 139 13 Glaurung 1.2.1 Crusader 32-bit 2765 +41 -41 52.2% -20.2 208 Zap!Chess Reykjavik 32-bit 2765 +35 -35 37.8% +78.5 255 14 CM10th Paralyse 2764 +25 -25 42.6% +49.3 535 Ktulu 7.5 2762 +16 -16 42.9% +48.5 1421 Glaurung 1.2.1 32-bit 2761 +17 -17 45.5% +31.5 1144 CM9000 Enforcer 2760 +33 -33 49.0% +4.4 291 CM10th Xperience 2759 +16 -16 42.0% +54.5 1404 LoopList 6.00 2759 +43 -43 54.3% -43.6 184 15 Smarthink 1.00 32-bit 2756 +16 -16 43.5% +42.9 1304 Spike 1.0a 2752 +19 -19 43.9% +41.2 871 CM10th Magic II 2746 +17 -18 40.6% +61.4 1037 16 Scorpio 1.8 2746 +19 -19 47.9% +14.7 877 CM10th Lazarus 2743 +41 -41 49.8% +1.4 200 CM10th Ogre 2743 +35 -35 47.1% +18.0 258 17 Slow Chess Blitz WV2.1 2741 +16 -16 48.8% +8.5 1250 CM10th Warrior 2738 +38 -38 46.3% +28.0 219 Ktulu 7.1 2738 +19 -19 40.9% +62.9 949 Chess Tiger 15 Gambit 2735 +31 -31 45.2% +37.1 347 CM10th Default 2733 +17 -18 40.0% +68.0 1089 Chess Tiger 15 2733 +17 -17 41.3% +59.9 1129 CM10th Pestilence 2729 +35 -35 49.4% +2.5 266 Glaurung 1.1 32-bit 2729 +38 -39 45.9% +31.4 231 18 WildCat 7 2727 +19 -19 45.1% +33.9 895 Glaurung 1.0.2 32-bit 2724 +18 -18 42.3% +56.0 1072 19 Gandalf 6 2723 +16 -16 40.0% +68.7 1411 20-21 Pro Deo 1.2 2722 +22 -22 49.7% -1.5 712 20-21 Ruffian 2.1.0 2722 +26 -26 43.5% +50.1 509 List 5.12 2720 +20 -20 46.4% +23.3 858 Hiarcs 9 2717 +20 -21 37.8% +82.1 784 Chess Tiger 2004 2716 +22 -22 40.4% +61.9 653 Delfi 5.1 2713 +236 -212 70.0% -108.8 5 Ruffian 1.0.5 2713 +21 -21 45.4% +29.7 732 Scorpio 1.7 2711 +36 -36 54.5% -32.1 245 Slow Chess Blitz WV2 2711 +19 -19 40.1% +69.0 950 22 Pharaon 3.5.1 2707 +18 -18 43.6% +44.9 998 23 Delfi 5.0 2703 +20 -20 47.5% +15.5 775 Naum 1.91 32-bit 2703 +18 -18 39.3% +72.5 1042 24 Aristarch 4.50 2699 +16 -16 42.7% +52.0 1335 WildCat 6 2697 +22 -22 47.1% +18.6 680 Delfi 4.6 2696 +31 -31 50.0% -2.3 334 25 Pseudo 0.7c 2693 +21 -21 49.5% +2.7 707 Pro Deo 1.1 2690 +32 -32 45.7% +30.9 323 Crafty 21.5 32-bit 2689 +105 -100 63.6% -90.7 33 Zappa 1.1 32-bit 2686 +40 -40 47.6% +18.3 207

by **Guenther Simon** » 24 Feb 2007, 11:32

Graham Banks wrote:
Guenther Simon wrote:I know no list in which Zappa on a single machine is anywhere
near 2700 in relation to what you list here.

Guenther

The key word was 'in relation'. Everyone can create a list,
where Zappa 1.1 has 7000 rating points whatever it means.
The ratings at CCRL are somehow inflated against all other
rating lists by 60-90 points. Obviously you scale different than
most others, may be also a consequence of yet much too less
entries below '2600' in your list?
In the real world Kiwi is not considered having 2567 rating
points anyway comparable to something like Elo or consensus
by other testers for the last years.
You cut off the part which could be compared with Tonys'...
(BTW you should consider testing FrancescaMad, it seems to miss
there?)

Code: Select all: Zappa 1.1 32-bit 2686 +40 −40 47.6% +18.3 32.4% 207 Arasan 9.5 2560 +43 −44 47.1% +19.6 24.3% 189 DanaSah 2.85 2536 +52 −52 49.6% +7.6 29.8% 124

Code: Select all: 1 Zappa 1.1 : 2600 36 35 360 71.5 % 2440 16.4 % 2 Arasan 9.5 : 2558 35 34 360 66.0 % 2443 15.3 % 3 Francesca MAD 0.13 : 2550 34 33 360 64.9 % 2444 19.2 % 4 Danasah 2.85 : 2498 32 32 360 57.1 % 2448 20.8 %

Regards,
Guenther

by **Graham Banks** » 24 Feb 2007, 19:14

Guenther Simon wrote:(BTW you should consider testing FrancescaMad, it seems to miss
there?)
Guenther

Hi Guenther,

we can't test Francesca, Bruja or Tytan because their internal opening books can't be disabled.
We don't allow engines to use their own books.

We can't test RomiChess either because learning can't be disabled (although Michael has said he'll make turning learning off an option at some stage).

Regards, Graham.

by **Tony Thomas** » 25 Feb 2007, 07:48

According to Leo's rating list Zappa(take off 40 points since he is using 64bit) is a 2700 engine, when I said add 100 points, I said some engines and I was refering mainly to engines towards the end of the list. There rating is rather skewed. He didnt test the same versions of other engines so...

Code: Select all: 17 Zappa 1.1 x64 2714 42 42 184 51% 2711 34% 47 FranMAD 0.13 2567 51 51 124 60% 2502 35% 70 Djinn 0.925x 2476 62 60 92 66% 2364 30% 76 Phalanx XXII 2455 18 18 1084 48% 2467 25%

by **Tony Thomas** » 27 Feb 2007, 06:30

Graham Banks wrote:
Guenther Simon wrote:(BTW you should consider testing FrancescaMad, it seems to miss
there?)
Guenther

Hi Guenther,

we can't test Francesca, Bruja or Tytan because their internal opening books can't be disabled.
We don't allow engines to use their own books.

We can't test RomiChess either because learning can't be disabled (although Michael has said he'll make turning learning off an option at some stage).

Regards, Graham.

It seems that you are not interested in work arounds. You can just delete the learnfile of Romichess after each 2 games. I know that is something hard to do, if I had more memory I would have helped you out with getting around two or three hundred games of Romi. Also, why not test Bruja, Tytan and Francesca and add on the name tag to specify that they are using own book? Also, when using a GUI book, the role of the own book is very small because none of those programs got a big book.

by **Graham Banks** » 27 Feb 2007, 19:01

Tony Thomas wrote:It seems that you are not interested in work arounds. You can just delete the learnfile of Romichess after each 2 games. I know that is something hard to do, if I had more memory I would have helped you out with getting around two or three hundred games of Romi. Also, why not test Bruja, Tytan and Francesca and add on the name tag to specify that they are using own book? Also, when using a GUI book, the role of the own book is very small because none of those programs got a big book.

Sorry - no compromises.
We test with learning off and with no engine using its own book - no exceptions.

by **Vladimir Medvedev** » 21 May 2007, 09:49

Version 5.3
20-May-2007
-----------

(chg) Using hash eval from previous iteration for pruning
(chg) Small optimization in attack detector
(chg) Small optimization in PV handling
(fix) Fixed crash on epd-files ended with empty lines

http://greko.chess.googlepages.com/greko-5.3-win.zip
http://greko.chess.googlepages.com/greko-5.3-linux.zip
http://greko.chess.googlepages.com/greko-5.3-source.zip

Enjoy!

by **Volker Pittlik** » 21 May 2007, 11:38

Vladimir Medvedev wrote:Version 5.3
...
greko-5.3-linux.zip
...
greko-5.3-source.zip

Thank you Vladimir!

I just compared the time neede for "perft 5" in UVI mode here on my Core2Duo E6300:

Linux executable delivered: 1.34 seconds
Own compilation using the Makefile delivered: 0.92 seconds
Compiled with -O3 option: 0.78 seconds.

Maybe others can make some tests too.

Volker

by **Vladimir Medvedev** » 21 May 2007, 12:13

Thank you, Volker!

I use gcc 3.3.3 under SuSE Linux. But I never test or profile GreKo under Linux, my development platform is Windows.
Can I use your Linux compilation as "official" one and put it on GreKo's page?

by **Volker Pittlik** » 21 May 2007, 12:41

Vladimir Medvedev wrote:Thank you, Volker!

I use gcc 3.3.3 under SuSE Linux. But I never test or profile GreKo under Linux, my development platform is Windows.
Can I use your Linux compilation as "official" one and put it on GreKo's page?

Sure here it is:http://www.vpittlik.org/temp/greko-5.3.bz2 . No guarantee of any kind of course! I'm also running SuSE but gcc version is 4.1.2.

by **Vladimir Medvedev** » 21 May 2007, 13:56

Thank you Volker for your Linux version. On my Linux box it works +24% faster than original (on perft 6 task). And version compiled by gcc 3.3.3 with -O3 surprisingly is -15% slower than original.

Winboard Forum

GreKo 5.3

GreKo 5.3

Re: GreKo 5.2

Re: GreKo 5.2

Re: GreKo 5.2

Re: GreKo 5.2

Re: GreKo 5.2

Re: GreKo 5.2

Re: GreKo 5.2

Re: GreKo 5.2

Re: GreKo 5.2

Re: GreKo 5.2

Re: GreKo 5.2

Re: GreKo 5.2

Re: GreKo 5.2

Re: GreKo 5.2

Re: GreKo 5.2.5

Re: GreKo 5.2.5

Re: GreKo 5.2.5

Re: GreKo 5.2.5

Re: GreKo 5.2.5

Who is online